Figures (2)  Tables (2)
    • Figure 1. 

      Natural product target identification based on protein libraries. (a) Thermal shift assay (TSA-CETSA-TPP), (b) drug affinity responsive target stability (DARTS), (c) stability of proteins from rates of oxidation (SPROX), and (d) target-responsive accessibility profiling (TRAP).

    • Figure 2. 

      Natural product target identification based on in-situ models. (a) Copper-catalyzed click chemistry (CUAAC), (b) strain-promoted azide-alkyne cycloaddition (SPAAC), (c) inverse electron-demand Diels-Alder reaction (iEDDA), (d) photoaffinity labeling (PAL), (e) degradation-based protein profiling (DBPP), (f) μMap, and (g) pup-on-target for small molecule target identification technology (POST-IT).

    • Natural product Activity Target Method used Mechanisms of action Ref.
      Acacetin Lipid metabolism regulation LAMTOR1 DARTS Modulates autophagy via the LAMTOR1-MTORC1-AMPK signaling pathway. Alleviates metabolic dysfunction-associated fatty liver disease [31]
      Artemisinin Against malaria Multiple parasite proteins SPAAC Targets across glycolysis, hemoglobin degradation, antioxidant defense, and protein synthesis pathways [86]
      Artemisinin Against malaria Targets across different stages of the parasite PAL Interferes with the protein synthesis, glycolysis, and oxidative homeostasis pathways of parasites [99]
      Artemisinin derivatives Antitumor PCLAF DBPP Degrades PCLAF and activates the p21/Rb axis [102]
      α-Mangostin Antitumor RTN4/Nogo Protein microarray Induces proteasome degradation of RTN4 through recruiting E3 ligase UBR5 and enhances the pyroptosis phenotype [72]
      Bufalin Antitumor ERα (ESR1) AI + Docking + MD simulation Enhance the interaction between ERα and the E3 ligase STUB1, thereby promoting proteasomal degradation of ERα [111]
      Celastrol Anti-inflammation CAP1 TRAP Ameliorates metabolic syndrome via cAMP-PKA-NF-κB pathway [62]
      Celastrol Antitumor, anti- inflammation, alleviating of metabolic disorders Multi-targets including CHK1, OGA, ERCC6L, IKKβ, PI3Kα, and CIP2A DBPP Inhibits PI3K/AKT pathway, NF-κB pathway, and induces G2/M phase arrest [100]
      Cycloastragenol Antitumor Cathepsin B TRAP Enhances CD8 T cell-mediated antitumor immunity [63]
      12-deoxyphorbol 13-palmitate Anti-liver fibrosis APOL2 PAL Disrupts APOL2-SERCA2-PERK-HES1 signaling and alleviates fibrosis [97]
      2,4-dihydroxybenzaldehyde Alleviation of acute
      kidney injury
      KEAP1 HiMAP-seq Inhibits KEAP1 and downstream and NRF2-GPX4 signaling axis [106]
      Eupalinolide B Anti-neuroinflammation USP7 Protein microarray Inhibits USP7 to cause a ubiquitination-dependent degradation of KEAP1 and further induces an NRF2-dependent transcription activation of anti-neuroinflammation genes [71]
      Evodiamine Antitumor REXO4 DBPP Degrades REXO4 to induce cell death through ROS [101]
      Forskolin Osteoporosis TGM2 CuAAC Allostericlly activates TGM2 to improve mitochondrial dynamics and ATP production for osteoblast differentiation [80]
      Fructose-1,6-bisphosphate (FBP) Signaling molecule ALDH2 PAL Inhibits ALDH2 activity and results in cellular ROS upregulation accompanied by mitochondrial fragmentation [96]
      Gambogic acid Antitumor CNPY3 TPP Facilitates the recruitment of SIRT1 to remove lysine lactylation of CNPY3 and disrupts lysosomal proteins [27]
      Geldanamycin Antitumor HSP90 SPROX Binds to HSP90's N-terminal ATP binding domain and its ATPase activity [43]
      Grincamycin B Antitumor IDH1 Protein microarray Inhibits IDH1 to disrupt the cellular redox balance and 2-oxoglutarate homeostasis, thereby triggers ROS accumulation and eventually causes cell death [69]
      Halorotetin B Antitumor UBE2C PELSA Induces M phase cell cycle arrest [65]
      Hyperforin Anti-obesity DLAT LiP-SMap Promotes thermogenesis by stimulating AMPK and PGC-1α via a Ucp1-dependent pathway [56]
      Kurarinone Alleviation of Parkinson's disease sEH SIP + TPP Suppresses sEH to stabilize the level of epoxyeicosatrienoic acids and inhibits GSK3β [47]
      Ligustroflavone Antitumor BRD4 HiMAP-seq Inhibits BRD4 and down-regulates MYC, NFKB1 and TP53 [106]
      Manassantin A Antitumor 28 targets (e.g., filamin A and elongation factor 1α) SPROX Induces conformational change in filamin A to interrupt filamin A-HIF1α interaction, thereby inhibits HIF1α [42]
      Matrine Alleviation of spinal cord injury HSP90 DARTS Enhances the chaperon activity of HSP90, leading to axonal growth [29]
      Naamidine J Alleviation of acute lung injury CSE1L CuAAC Inhibits SP1 nuclear translocation and suppresses macrophage inflammation [79]
      Neoeriocitrin Osteogenesis and bone regeneration BECLIN1 TPP Stabilizes BECLIN1 by inhibiting ubiquitination-mediated degradation, thereby increasing autophagy [26]
      PF403 Antitumor NAMPT TPP Inhibits the catalytic activity of NAMPT, leading to a decrease in the NAD+ concentration [25]
      Shikonin Antitumor IKKβ/NEMO complex TPP + SIP Destabilizes the IKKβ/NEMO complex to suppresses NF-κB signaling and impairs cell proliferation [48]
      Silibinin Hepatoprotection ACSL4 TRAP Inhibits ACSL4 enzymatic activity, thereby mitigating the ACSL4-mediated ferroptosis [61]
      Tetrandrine Antivirus and antitumor LIMP-2 PAL Binds LIMP-2's ectodomain to inhibit lysosomal cholesterol and sphingosine transport [98]
      Walrobsin A Alleviation of acute kidney injury GPR75 DARTS Inhibits GPR75 to alleviate macrophage oxidative stress and the inflammatory microenvironment [30]

      Table 1. 

      Representative example of natural product target Identification.

    • Method Principle Key features Advantages Limitations
      TSA/CETSA/TPP Ligand binding alters protein thermal stability
      Tm shift detected via fluorescence (TSA), immunoblot (CETSA), or MS (TPP)
      – Label-free
      – TSA: purified proteins
      – CETSA: cells/lysates
      – TPP: proteome-wide MS
      – No modification
      – CETSA: live-cell compatible
      – TPP: high-throughput, unbiased
      – TSA: pure protein only
      – CETSA: low throughput
      – TPP: costly, long runtime, membrane protein under-representation
      DARTS Drug binding confers resistance to proteolytic digestion – Label-free
      – SDS-PAGE/MS readout
      – No probe synthesis
      – Works with crude extracts
      – Detects low-affinity interactions
      – False positives
      – Low-abundance proteins missed
      – Protease choice critical
      SPROX Ligand stabilizes protein, reducing methionine oxidation rates; quantitated by MS – Chemical denaturation gradient
      – MS detection of oxidized/non-oxidized Met peptides
      – Quantitative affinity data – Requires Met-containing peptides
      – No binding site information
      SIP Organic solvents precipitate free proteins; ligand-bound proteins remain soluble. – Mixed organic solvent
      – MS readout
      – Good proteome coverage
      – No probe modification
      In vitro only
      – Organic solvent may disrupt some interactions
      LiP-MS Ligand-induced conformational changes alter protease accessibility; MS quantifies differential peptides. – Two-step proteolysis
      – Peptide-level resolution
      – Binding site mapping (~12 aa)
      – Proteome-wide
      – Near-physiological conditions
      – Requires high sequence coverage
      – Cannot localize distal allosteric sites
      TRAP Ligand binding changes lysine accessibility to reductive dimethylation – Isotope-coded formaldehyde labeling
      – MS readout
      No protein modification – Requires reactive Lys near binding site
      PELSA Local stabilization upon ligand binding hinders tryptic cleavage; reduced peptide abundance pinpoints binding region – High-concentration trypsin digestion – No modification
      Direct binding site information
      – High sensitivity
      – May need optimization for each protein
      – Sequence coverage dependent
      Protein microarrays Immobilized recombinant proteins probed with labeled or unlabeled NP – High-density chip
      – Fluorescence/MSI/DESI-MS readout
      – Ultra-high throughput
      – Parallel screening of thousands of proteins
      – Improper protein folding and truncated fragments
      – Non-physiological in vitro environment
      – High experimental cost
      CuAAC Azide/alkyne-modified NP; Cu(I) catalyzes cycloaddition with biotin tag for pulldown – Bioorthogonal probe
      – Affinity enrichment
      – Fast, specific, sensitive – Copper cytotoxicity
      – Chelation inhibits catalysis
      SPAAC Strain-promoted cycloaddition; no catalyst – Copper-free
      – Ring-strain driven
      – No metal toxicity
      – Live-cell compatible
      – Slower kinetics (~0.1 M−1·s−1)
      – Larger functional groups
      iEDDA Tetrazine (Tz) + trans-cyclooctene (TCO) inverse electron-demand Diels-Alder – Extremely fast – Ultrafast kinetics
      Bioorthogonal
      – Bulky Tz/TCO may alter NP properties
      PAL UV irradiation activates photoreactive group; covalent crosslinking to target – Probe: NP + photoreporter + tag
      – UV-induced carbene/nitrene
      – Captures transient/weak interactions
      – Membrane and low-abundance proteins
      – Probe synthesis may alter activity
      – UV damage
      DBPP NP-PROTAC conjugate recruits E3 ligase; target degradation – Ternary complex formation
      – Quantitative proteomics
      – Detects moderate/weak binders
      – Direct pulldown evidence
      – Complex design
      – Costly, platform-dependent
      μMap Visible-light photocatalyst (Ir) conjugated to NP; Dexter energy transfer activates diazirine probe nearby – Photocatalytic proximity labeling
      – 450 nm blue light
      – No UV damage
      - High membrane protein enrichment
      – Poor tissue penetration (blue light)
      POST-IT HaloTag-PafA fusion recruited by HTL-NP; pupylation of proximal lysines in live cells/animals – Prokaryotic pupylation system
      – Non-diffusive labeling
      – Live-cell and in vivo
      – Preserves transient interactions
      – Requires genetic fusion
      – Engineering of PafA ligase
      HiMAP-seq Multiplexed RNA-seq with UMIs; generates Chemical-Induced Gene Signatures (CIGS) for mechanism inference – Pooled sample barcoding
      – No mRNA purification
      – Ultra-high throughput
      – Cost-effective, reproducible
      – Indirect target inference
      – Requires validation
      AI/
      Bioinformatics
      Integrates multi-omics, network models, machine learning; predicts targets from chemical structure and expression data – Computational pipelines
      – Deep learning, network pharmacology
      – Accelerates hypothesis generation
      – Reveals polypharmacology
      – Prediction accuracy depends on training data
      – Requires experimental confirmation

      Table 2. 

      Comparative summary of natural product target identification methods.